Go Back   SEQanswers > Bioinformatics > Bioinformatics

Similar Threads
Thread Thread Starter Forum Replies Last Post
Qubit library quantitation variations Smriti Sample Prep / Library Generation 5 11-02-2013 10:57 AM
Sample/library prep of DNA and RNA in a metagenomic sample chrisaw01 Metagenomics 1 05-05-2011 02:59 PM
Variations off GATK JohnK Bioinformatics 2 02-17-2011 12:21 PM
Reproducible variations of accuracy Michael L. Altshuler Illumina/Solexa 0 10-09-2010 03:21 AM
Structural Variations sparks Bioinformatics 0 10-30-2008 03:16 PM

Thread Tools
Old 01-03-2011, 07:57 PM   #1
Location: Hong kong

Join Date: Jul 2010
Posts: 17
Default Identifying variations from one sample

Hi, can anyone suggest one good method to identify variations from one NGS sample? I have read some methods in published papers, it seems that lots of them consider the observed priori probability of mutation is the same with SNP, 0.001. However, I think the ratio of heterotype SNVs should be larger than that, probably near 0.5. (Maybe I am confused, 0.001 is very useful for novel mutations finding) Meanwhile, is there any good programs for somatic mutations discovering, when two samples are used (control+)? I have tried to identify them by the following procedure: a.Identify mutations in different samples respectively.b. Find the difference between them and name them somatic mutations. However, this could generate a big number of somatic mutations. I think it would be more reasonable if the number is a bit smaller. Besides, this procedure does not consider mutations that occur in both of the two samples, but vary in allele frequency. Very appreciated if you could give me some suggestions.

Last edited by dingxiaofan1; 01-03-2011 at 08:14 PM.
dingxiaofan1 is offline   Reply With Quote
Old 08-17-2011, 11:06 AM   #2
Location: spain

Join Date: Feb 2011
Posts: 60

I do not know if it can help to you, but it seems that the optimal pipeline for finding somatic mutations in tumor/normal samples with GATK is the following:

That is, to work with a single bam file for each tumor/normal pair throughout the workflow steps once their reads are mapped.

Actually, I'm also working on it, since once the normal/tumor mutations are detailed in the final single vcf file, I do not know which is the most approppriate way of declaring them as either somatic or germline.
david.tamborero is offline   Reply With Quote
Old 08-17-2011, 02:59 PM   #3
Registered Vendor
Location: Seattle, WA

Join Date: Nov 2009
Posts: 6


Can you explain exactly what you mean when you say "variations from one NGS sample"? Do you mean between that sample and a reference sequence or the between the two haplotypes in that single individual, which would be the number of heterozygous SNVs? I think the .001 probability is what you want to use in either case. What variant calling procedure are you using?

I think the procedure you describe is the most common way of looking for somatic mutations- call variants in both control and non-control samples and find the differences. Depending on your data and your variant calling method it should be possible to detect SNVs mutations that occur in both samples,but are heterozygous in one and homozygous in the other.

dreesbl is offline   Reply With Quote

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

All times are GMT -8. The time now is 05:25 PM.

Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2021, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO